We study the influence of a single Feshbach resonance on the near-threshold bound and continuum states in interatomic potentials with attractive tails. We present an energy-independent parametrization of the Feshbach resonance to reveal its influence on the bound state spectrum as well as on the scattering properties. The applicability of this parametrization is confirmed by comparison with numerical coupled-channel calculations. A formula for the scattering length is derived, which---in contrast to previous parametrizations---transparently exposes the dependence on the characteristic quantities of the uncoupled channels and on channel-coupling effects and is valid for all deep potentials falling off faster than $\ensuremath{-}1/{r}^{3}$ asymptotically.
The interactions between atoms and molecules with each other and with surfaces are typically deep potential wells with attractive tails behaving asymptotically as an inverse power of the distance. In such potential wells, bound state energies
En are determined by a quantization function
F(E) according to ${n}_{\mathrm{th}}\ensuremath{-}n=F({E}_{n})$, and
F(E) is dominantly determined by the singular potential tail for near-threshold states. In this paper we formulate a general theory for the contribution
Ftail(E) of the singular potential tail to the quantization function. The general expression for
Ftail(E) contains two terms: a difference of action integrals and a difference of outer reflection phases, taken at threshold and at energy
E. Close to threshold,
E=0, strongly energy dependent and nonanalytic contributions of both terms cancel, so
Ftail(E) acquires a universal form determined by a threshold length and an effective length which is related to a subthreshold effective range. For a homogeneous potential tail proportional to $\ensuremath{-}1∕{r}^{\ensuremath{\alpha}}$, one universal expression for
Ftail(E) caters for all potential strengths. We give an explicit analytical expression for the important case $\ensuremath{\alpha}=6$ and present applications involving the derivation of atom-atom scattering lengths from the binding energies of high-lying bound states of the associated diatomic molecule. We also demonstrate how the dissociation energy of a diatomic molecule can be determined from spectroscopic energies of high-lying states, and we make a quantitative comparison with the performance of the LeRoy-Bernstein formula, which fails near threshold, because the strongly energy dependent and nonanalytic contribution from the action integrals is not, as it should be, compensated by terms coming from the corresponding energy dependence of the outer reflection phase.
For potential wells with long-range attractive tails proportional to $\ensuremath{-}1/{r}^{3}$, as occur in the resonant dipole-dipole interaction in homonuclear alkali-metal dimers, we present a highly accurate analytical expression for the tail contribution to the quantization function
F(E). This quantization function determines the near-threshold bound-state energies via the quantization rule ${n}_{\mathrm{th}}\ensuremath{-}n=F({E}_{n})$. The performance of the quantization function derived in this paper is demonstrated by applying it to a model Lennard-Jones potential and to vibrational bound-state spectra of sodium dimers (Na
2). These results are compared with those obtained via the semiclassical LeRoy-Bernstein formula which neglects quantum effects that are important in the near-threshold regime.
For potentials with attractive tails, as occur in typical atomic interactions, we present a simple formula for the
s-wave phase shift ${\ensuremath{\delta}}_{0}$. It exposes a universal dependence of ${\ensuremath{\delta}}_{0}(E)$ on the potential tail and the influence of effects specific to a given potential, which enter via the scattering length
a, or equivalently, the noninteger part ${\ensuremath{\Delta}}_{\mathrm{th}}$ of the threshold quantum number
nth. The formula accurately reproduces ${\ensuremath{\delta}}_{0}(E)$ from threshold up to the semiclassical regime, far beyond the validity of the effective-range expansion. We derive the tail functions occurring in the formula for ${\ensuremath{\delta}}_{0}(E)$ and demonstrate the validity of the formula for attractive potential tails proportional to
1/r6 or to
1/r4, and also for a mixed potential tail consisting of a
1/r4 term together with a non-negligible
1/r6 contribution.
We analyse the influence of the attractive tail of the p–H two-body potential on the energy progression of the near-threshold rotationless bound states of the H molecular ion. For the 1sσ g configuration we study different definitions of the reference potential describing the potential tail. In each case, the scattering length consistent with the energies of the highly excited vibrational states, as calculated by Hilico et al. [Eur. Phys. J. D 12, 449 (2000)], is found to lie in the range between −30.5 and −31 au. In the 2pσ u configuration, for which there is an extremely weakly bound v = 1 vibrational state [Carbonell et al., Europhys. Lett. 64, 316 (2003)], we obtain a scattering length near 762.8 au. In contrast to previously conveyed impressions, the weakly bound v = 1 state can be satisfactorily described as the excited state in a distant shallow potential well of the p–H two-body potential, provided that a sufficient number of terms is included in the representation of the large-distance behaviour of this potential.
We formulate the quantization rule ${n}_{\mathrm{th}}\ensuremath{-}n=F({E}_{n})$ for bound-state energies
En in a deep potential well with a long-range attractive tail for an extended range of near-threshold energies. This makes it possible to accurately read off the scattering length
a from the energy of a high-lying (not necessarily the highest-lying) bound state via $a=\overline{a}+b∕\mathrm{tan}[\ensuremath{\pi}F({E}_{n})]$. For a potential tail proportional to $\ensuremath{-}1∕{r}^{\ensuremath{\alpha}}$, $\ensuremath{\alpha}>3$, the constants
a and
b are known and an explicit and universally valid expression for the function
F(E) is given analytically.